- Market
- Saudi Arabia
- Written for
- Head of Digital, Commercial Director
- Decision
- Fix the technical structure that limits what the site can rank for
- Languages
- EN · AR
The structure
Where search visibility is engineered: taxonomy, canonicals, schema, crawl
Four layers, each an engineering deliverable with a testable output. Content strategy decides what the site says; these decide whether a search engine can act on it.
URL taxonomy
One URL per search intent, in a hierarchy that mirrors how buyers actually phrase their searches, with internal links that carry authority to the pages that sell.
What the crawler then sees: A site whose structure answers the query before the copy does, with no two pages competing for one search.
Canonicals and hreflang
Every Arabic–English pair declared as alternates of each other, one canonical per intent, emitted by the page templates so a new page cannot ship without them.
What the crawler then sees: The Arabic searcher lands on the Arabic page, and duplicate variants stop splitting one page’s authority into fractions.
Structured data
Machine-readable descriptions of the organisation, services, articles and FAQs that match the visible content exactly — never claims the page itself does not make.
What the crawler then sees: An unambiguous machine reading of each page, and eligibility for the richer result formats search engines reserve for it.
Crawl efficiency
Filters, parameters, pagination and archives brought under control, with segmented sitemaps that tell the crawler where the pages worth its time live.
What the crawler then sees: A crawl budget spent on pages that earn revenue instead of on infinite filter permutations.
The distinction
This is engineering work, and it is finished when the defects are gone
Most of what is sold as SEO is content marketing on a retainer — articles, links, monthly reports. Useful, sometimes; but it sits on top of the architecture, and when the architecture is broken it is spending poured into a cracked vessel. Two pages competing for one intent, an Arabic tree half-connected to its English counterpart, a crawler exhausting its visit on filter permutations: no article fixes these, because they are properties of the codebase, and they are corrected in the codebase — in templates, routing and build checks, by engineers.
The other difference is that engineering work ends. An architecture engagement produces a fixed list of defects, closes them, and installs the checks that stop them returning — a build that fails when a page ships without its canonical or its language counterpart. What remains ongoing afterwards is content, which is a different discipline with a different owner. We would rather hand over a sound structure and step back than convert a finite repair into an indefinite retainer.
The mechanism
Collapsing every variant URL into the one version worth a crawler’s time
A working site generates far more URLs than it has pages. The canonical map is the control layer that tells search engines which of them count.
Many URL sources — two language trees, filters, tracking variants, pagination — converge into a canonical map that produces a focused crawl, one result per intent per language, and a structure that survives growth.
Two language trees
Every page exists in Arabic and English, doubling the URL count before anything else does.
Filters and parameters
Sort orders, filters and query strings mint new URLs for the same content on every combination.
Tracking variants
Campaign tags append themselves to real pages and travel into indexes as if they were new ones.
Pagination and archives
Listing pages stretch a handful of items across long, thin chains of near-identical URLs.
What we do
The canonical map
One declared, authoritative URL per intent and per language, emitted by the templates and enforced by a build check — never annotated by hand across thousands of pages.
A crawl spent on real pages
The crawler’s finite visit lands on pages that can earn, not on variants of them.
One result per intent, per language
Authority consolidates on a single URL instead of splitting across its own duplicates.
A structure that survives growth
New sections inherit the rules automatically, so scale adds pages without adding chaos.
Our role
What the architecture engagement repairs and what your content team aims
- 01
The audit, with evidence
We list the structural defects — duplicate intents, broken language pairing, wasted crawl, missing or dishonest markup — each traced to what search engines actually fetched, not to a tool’s generic score.
- 02
The fixes, in the codebase
Taxonomy, canonicals, hreflang, structured data and crawl controls are corrected in templates and routing, so the fix applies to every page that exists and every page yet to be written.
- 03
The guardrails that outlast us
Build checks fail any release that breaks a canonical, orphans a language pair or ships markup that contradicts the page — so the structure holds without anyone having to remember it.
What the site says, to whom, and in which words is content strategy and stays with your team; our work ensures that nothing structural caps what that content can earn — and no one, including us, can honestly promise you a position.
Common questions
What buyers ask
Can anyone guarantee our site a first-page ranking?+
No — and a supplier who does is describing a promise the search engine has not delegated to them. Ranking is decided by algorithms neither party controls, against competitors neither party controls. What can honestly be committed to is the engineering: defects found, fixed and verified against what crawlers actually fetch, so nothing structural stands between your content and the results it can earn. We put the guarantee where the control is — in the state of the site — and we would treat a rankings guarantee from anyone as a reason to read the contract twice.
How should hreflang connect the Arabic and English versions of the same page?+
As a symmetrical pair: the Arabic page declares the English one as its alternate, the English page declares the Arabic one, each includes itself, and the URLs referenced are the canonical, absolute ones. Break any side of that — one direction missing, a redirect in the chain, a canonical pointing elsewhere — and search engines quietly fall back to guessing, which is how Arabic searchers end up on English pages. This is exactly why we emit the annotations from the page templates rather than maintaining them by hand: a symmetry that depends on human memory across thousands of pages will not stay symmetrical.
Why does search keep showing the wrong page from our site for queries we should win?+
Almost always because more than one of your pages answers the same intent, and the search engine made the choice you never made. When a service page, a blog post and a landing page all half-answer one query, the algorithm picks — and it often picks the page you least want a buyer to land on, while the votes split three ways. The architectural repair is to assign each intent exactly one owning URL, fold or canonicalise the rivals into it, and let internal links state the hierarchy. This is a structural decision expressed in the codebase, which is why it belongs to this work and not to a content calendar.
In this cluster
Request the audit
A technical audit that names the architectural defects hiding your pages from search, ordered by what fixing each one is likely to release.
Start the conversation
